| The 1W could be the correct AbsMax rating. It is a 100mm device with proportionally sized heat sink connections as the similar looking 5mm "0.5W" ones.
The LED marketing very frequently incorrectly use the AbsoluteMaximum rating as the operation characteristics, part of it is the fact,, that these AbsMAx are the only place in the datasheet, where you find the values like power and/or current figures.
The same way the fact, the lumen output is rated at given current does not mean such current is safe for long term use.
Because there are other figues as well, like maximum junction temperature, reliability data, thermal property parameters (thermal resistances from the virtual junction,...) and so on. So when you design something, your duty is to verify, than no single value listed in the limitting value list is exceeded even for very short time, including any possible component tolerance and external condition variations and the device operate under such condition, the reliability would be acceptable (life, degradation, failure rate,...).
This is the way, how such data are published in the complete electronic industry (rating for short term peaks, lifetime data for failure rate in ppm, parameter tolerances,...)
Tt differ from what the lighting industry use to publish (nominal voltage, median lifetime, no ratoing for excessive conditions at all, no tolerance figures,...).
The reason is, than the exact current/power the device could handle very strongly depend on the actual fixture design, so you can not put them into the datasheet (datasheet values should reflect only the component, nothing else, otherwise the document become an unusable mess). So it is the role of the product design engineer to evaluate the safe levels.
Here is, in very shortened and simplified form, how to ensure you won't overload the led's:
Get the figure of thermal resistance (Rthjc) between the junction and the cooling lung attachement point from the datasheet and the maximum operational (not abs-max; if not specified, it is AbsMax-25degC) junction temperature (Tjmax).
Now you should establish a maximum expected temperature on the cooling lungs (Tcmax; it is an initial guess, it would be subject of optimization later on). It would be quite above the ambient, ~80..90degC would be the first guess I would use on a simple PCB design
Now you could calculate the maximum power the LED could handle: Pmax=(Tjmax-Tcmax)/Rth
Now you compare it with the AbsMax power rating, of course the maximum load value you need would be the lower one from these two, but you will see, than usually the Pmax you have calculated would already be not much more than half of the AbsMax rating.
Then you should assemble one prototype and drive it in the calculated manner. During that you should monitor the temperature of the LED luing and other critical points (tabs of power transistors, transformers, all the components where you expect the heat to be generated,...). Make sure the enclosures are closed as they would be in normal operation, as they affect the cooling.
Now measure the margin, how far you are from the Tcmax on each of the ctritical component (the Tcmax is individually obtained for each one, in a similar way as for the LED's).
Add this difference to the actual room temperature and so you would get the maximum ambient temperature to operate this. My guess, you won't end up much above 30..40degC...
If this is too low, either improve the heatsink, or select higher value of the Tcmax and reduce the drive accordingly.
Now if you fit the AbsMax power rating figures into these equations, you end up (again my guess) in maximum ambient temperature around -20..-30degC... Well, pretty cold limit for the upper temperature limit...
|
